Adversarial AI Debate

This skill transforms Claude into a rigorous thinking partner by facilitating structured debates with external models like Gemini, Codex, and Qwen. Rather than Claude debating itself, it leverages the unique training and biases of different LLMs to provide genuine pushback, surface conflicting assumptions, and document complex tradeoffs. It is particularly useful for high-stakes architecture decisions, strategy planning, or any scenario where a 'yes-man' response would be detrimental to the final outcome, resulting in a synthesized consensus or a detailed tradeoff document.

主要功能

使用场景